Ekaterina Hristova Taralova
ekaterina@cmu.edu



In my thesis, I augment machine learning algorithms (e.g., Bag of Words, clustering, hierarchical video segmentation) with domain-specific features (e.g. local, global, spatial, temporal) to build robust and semantically meaningful representations for videos.

Curriculum vitae


In December 2014 I will join the Rafael Yuste neuroscience lab at Columbia University as a postdoc fellow.
Intern at Facebook's Crowdsourcing team, working with Venky Iyer, August 2014 - December 2014.
Intern at Microsoft Research, working with Eric Horvitz, May 2014 - August 2014.

Research

Motion Words for Videos,
Ekaterina Taralova, Fernando De la Torre Frade, and Martial Hebert,
European Conference on Computer Vision (ECCV), September, 2014. [PDF]
Feature Quantization and Pooling for Videos,
Ekaterina Taralova
PhD thesis, April 2014, PDF will be available August, 2014.
Structured Models for Videos,
Ekaterina Taralova
Thesis proposal, August 2013, PDF available upon request.
Source Constrained Clustering,
Ekaterina Taralova, Fernando De la Torre Frade, and Martial Hebert,
International Conference on Computer Vision (ICCV), November, 2011. [PDF] [citation]
Modelling and visualizing morphology in the fungus Alternaria,
Ekaterina H. Taralova, Joseph Schlecht, Kobus Barnard, and Barry M. Pryor,
Fungal Biology, Vol. 115, pp 1163-1173, 2011 [PDF] [citation ]
Temporal Segmentation and Activity Classification from First-person Sensing,
Ekaterina H. Spriggs, Fernando De la Torre Frade, and Martial Hebert,
IEEE Workshop on Egocentric Vision, CVPR 2009, June, 2009. [PDF] [citation]
Inferring Grammar-based Structure Models from 3D Microscopy,
Schlecht, Joseph; Barnard, Kobus; Spriggs, Ekaterina; Pryor, Barry,
Computer Vision and Pattern Recognition (CVPR), 2007 [PDF] [citation]
Predicting emotional experience from autonomic physiology using machine learning methods,
Ekaterina Spriggs, Emily A Butler, Kobus Barnard, Frank H Wilheim, James J Gross,
Psychophysiological Research, 2007, poster only

About

I joined the Computer Science PhD program at Carnegie Mellon University in Fall 2008. I am advised by Dr. Martial Hebert and Dr. Fernando De la Torre both from the Robotics Institute.

I am interested in sensing the environment through computer vision and machine learning. In specific, my PhD thesis is on automatic temporal segmentation and classification of actions in videos.

I am also very passionate about educational outreach in science at the middle and high school levels.

I received my B.S. in Computer Science and Mathematics from University of Arizona in 2007, where I was advised by Dr. Kobus Barnard, and worked closely with Dr. Barry Pryor and Dr. Emily Butler.


Contact

Email: ekaterina AT cmu.edu
Mailing address:

Ekaterina H Taralova
CMU 6105 GHC
Computer Science Department
5000 Forbes Ave
Pittsburgh, PA 15213


Projects and service

Open House Committee and DRC Committee at CMU
NACLO: North American Computational Linguistics Olympiad Program Committee
All for Education